1

Annex E

Results of the questionnaires on operationalizationand summary of the structure and principles of standardizationof EPIC-SOFT

1

Annex E-1

EFCOSUM; WORKING GROUP 4

RESULTS OF THE FIRST QUESTIONNAIRE ONOPERATIONALIZATION OF A EUROPEAN FOOD CONSUMPTIONSURVEY

Participating countries (abbreviated in the tables below):AU = Austria GR = Greece PL = PolandBE = Belgium HU = Hungary PT = PortugalCR = Croatia IC = Iceland SK = Slovac RepublicCZ = Czech Republic IR = Ireland SP = SpainDK = Denmark IT = Italy SW = SwedenFI = Finland LI = Lithuania SZ = SwitzerlandFR = France NL = Netherlands UK = United KingdomGE = Germany NO = Norway

RESPONSIBLE INVESTIGATOR(S):

1. Principle investigator for a nationwide dietary survey?- Institute of Nutritional Sciences (AU)- At this moment there is no official Food Research Institute. For health monitoring: Scientific

Institute of Public Health – Louis Pasteur (IPH). In future food consumption survey probably alsocoordinated by this Institute. Possible alternative: ad hoc Task Force of experts in nutrtition fromdifferent Belgian Universities. (BE)

- Croatian institute of Public Health (CR)- National Institute of Public Health in Prague (CZ)- Danish Veterinary and Food Administration (DVFA); Institute of Food Research and Nutrition

(DK)- The National Public Health Institute (=KTL) (FI)- AFSSA / InVS (FR)- Robert Koch Institute (GE)- Dept. of Hygiene and Epidemiology, School of Medicine, University of Athens (GR)- National Institute of Food Hygiene and Nutrition (HU)- Icelandic Nutrition Council (IC)- Irish Universities Nutrition Alliance (IR)- Istituto Nazionale di Ricerca per gli Alimenti e la Nutrizione (INRAN) (IT)- National Nutrition Centre (LI)- TNO (NL)- National Council on Nutrition and Physical Activity in cooperation with Institute for Nutrition

Reseach, University of Oslo (NO)- National Food and Nutrition Institute with cooperation with other institutions (PL)- National Institute of Health (PT)- Medical Faculty of Comenius University in collaboration with the State Health Institute (SK)- Dept. of Nutrition. UCM Institute of Statistics (SP)- Swedish National Food Administration (NFA) alone or in collaboration with universities (SW)- Federal Health Institute (SZ)- UK Food Standards Agency would be responsible for commissioning such a survey (it would be

put out to tender) (UK)

2

3

2. Market research company preferred to carry out (parts of) such a survey?Answers N Countries1 yes2 noDependent on …Not decided yetNot answered

109121

AU,BE,DK,GE,GR,LI,NL,SK,SP,SWCR,CZ,FI,HU,IC,IR,NO,PL,PTFRIT,SZUK

Remarks:- Dependent on available budget. Market Research companies are usually expensive (and can be

uncooperative to some extent). (AU)- In collaboration with regional Institutes of Public Health (also engaging students of medicine,

nutrition and dieticians; professionals with experience in nutrition. (CR)- Very expensive (CZ,PL)- Normal procedure in Denmark (DK)- National FINRISK Study (national health monitoring) is carried out by the National Public Health

Institute, and the national dietary survey is carried out in a subsample of this study. (FI)- Depends on size of the sample and time available to prepare the fieldwork. Large sample (>5000)

and 1.5 year to prepare the fieldwork à National Institute of Statistics; Else à Market ResearchCompany (FR)

- A market research company would be useful for sampling and recruiting, but very expensive. Alsoit is preferable to have nutritionists recruiting subjects for optimum response (IR)

- Our Institute has the necessary skill for planning such a survey but not the number of fieldworkers. So far INRAN has not implemented a stable monitoring system (ad hoc financed). Itcould be resolved if our governmental organisations will be directly involved, to warrant thenecessary budget on a regular base. (IT)

- Good experience with 3 national surveys (NL)- Five year contract with the Institute for Nutrition Research regarding national dietary surveys. But

we are open to use market research companies to carry out parts of a survey (NO)- No experience with market research companies, but the National Institute of Statistics can perform

parts of the study (PT)- It appears to be realistic that under some conditions the institutions mentioned under 1. could

undertake the responsibility for a nationwide survey (SK)- But National Institute of Statistics had the census available (SP)- Previous nationwide surveys were carried out in collaboration with Statistics Sweden, which have

been responsible for sampling and fieldwork. In principle, a larger dietary survey has to be putunder public tender (partly or as a whole), unless carried out by NFA alone or in collaborationwith university departments. (SW)

4

SAMPLING:

3. Census available for representative sampling of adults on an individual basis?Answers N Countries1 yes

2 no

18

5

AU,BE,CR,CZ,DK,FI,FR,GE,HU,IC,IT,LI,NL,NO,PL,PT,SP,SWGR,IR,SK,SZ,UK

Remarks:- But these data are very sensitively discussed in the public in terms of data protection (AU)- But also election lists available (CR)- New census expected next year (CZ)- Civil Registration System (CRS) usually used for random sampling for major food intake studies.

The civil register contains current information about each citizen, including personal ID + formeraddresses, names, civil status and citizenship. (DK)

- A population Information System (FI)- Census gives sample of households. You have to choose randomly an individual in each

household (FR)- There is census, but not available for representative sampling of adults (access to names is not

allowed). Electoral register is used. (IR)- Statistics Norway draw the representative samples (NO)- Census was performed in 2001, but neither names nor ID codes were collected. Electoral lists

appear to be a better solution (SK)- There is a census, but it cannot be used for sampling for confidentiality reasons. Postcode

Address File is preferred (UK)

4. Feasible to use census for sampling? Institutions or market research companies that canprovide population samples on request?

Answers N Countries1 yes

2 noNo/yes

18

41

AU,BE,CR,CZ,DK,FI,FR,GE,HU,IC,IT,LI,NL,NO,PL,PT,SW,SZGR,IR,SP,UKSK

Remarks:- Dependent on available budget. Market Research companies are usually expensive (and can be

uncooperative to some extent). (AU)- Expected that co-operation will be possible with the Czech Statistical Office (CZ)- Population Register Centre (FI)- National Institute of Statistics (INSEE) can give household samples, but it takes a long time (FR)- Central Statistical Office (HU)- University of Iceland (IC)- See remarks question 3 (IR)- Limits of budgets are always to be taken into account (IT)- If paid (NL)- ‘PESEL’ office (PL)- National Institute of Statistics can provide population samples on request (PT)- Census probably not. Market research companies could provide population sample, however, they

are expensive (SK)- It is necessary to ask to go to judicial permission (SP)- Census preferred; some market research institutes may use other techniques (SW)- See remark question 3 (UK)

5

5. Feasible to sample children and adolescents through schools?Answers N Countries1 yes

2 no

21

2

AU,BE,CR,CZ,DK,GE,GR,HU,IC,IR,IT,LI,NL,NO,PL,PT,SK,SP,SW,SZ,UKFI,FR

If yes, type of schools:- 6-18 y: elementary, primary and secondary schools; 3-6 y: preschools (with some limitations due

to private preschools) (AU)- all ages: all types of schools (private and public) (BE)- 6-14 y: elementary schools; 15-18 y: high schools (CR)- 6-15 y: basic schools; 16-19 y: secondary schools; 20-25 y: Universities (CZ)- 7-16 y: primary school. But usually CRS sampling will be used (see Q.3) (DK)- 8-14 y; 95% coverage (GE)- Only for public, not for private schools; 6-12 y: elementary school; 13-15 y: high school; 16-18 y:

Lyceum (GR)- 15-18 y (HU)- 6-16 y; elementary school (IC)- 1-11 y: primary school; 12-18 y: secondary school (IR)- 6-10 y: primary school; 11-13 y: junior high school; 14-18 y: high school (in future it will depend

on the type of reform Government intend to implement (IT)- 6-10 y: primary school; 10-16 y: basic school: 16-18 y: secondary school (LI)- 0-18 y: in cooperation with National Health Surveillance (NL)- 6-14 y: school; 16-18 y: coverage not so good; 18 y: about 85% attend secondary schools; > 16 y

are usually included in national dietary or market surveys (NO)- 7-12 y: elementary schools; 13-18 y: secondary and vocational schools (PL)- 7-15 y: basic school; 16-18: secondary schools (PT)- 3-6 y: kindergartens; 6-18 y: primary and secondary schools (SK)- 3-17 y: public and private schools (SP)- for all levels/school types (SW,SZ)If no, what type of other sampling?- Through Population Register Centre. Sampling through schools is a tedious process, needs

permission of schools and municipalities. (FI)- Household samples (FR)

6

RECRUITMENT AND HANDLING OF NON-RESPONSE:

6. Type of institute or organisation likely to positively influence response rate for firstapproach of subject?

- Institute of Nutritional Sciences + non-governmental organisations as the Austrian Cancer Aid(AU)

- Ministry of Health / High Council on Hygiene; Research institutes (BE)- Croatian Institute of Public Health and Network of Country Institutes of Public Health; Ministry of

Health (CR)- Combination of international and local institutes (CZ)- Government institutions (DK)- Best: National Public Health Institute. Second best: Statistics Finland. Both have better image than

market research companies. Also Universities have a positive image (FI)- INSEE = National Institute of Statistics (FR)- Federal Health Research Institute, like DIfE or RKI (GE)- Institute from the Academic Environment (e.g. University) (GR)- National Public Health and Medical Officer Service; National Health Promotion Institute (HU)- University of Iceland or Nutrition Council or Office of the Surgeon General (IC)- Nutrition Universities of IUNA, Dept. of Health, the health of the individuals themselves, personal

health (IR)- INRAN (see 1) or National Institute of Statistics (ISTAT) (IT)- National Nutrition Centre + EU Institutions (LI)- Ministry of Public Health, Welfare and Sports (NL)- Local Institutions as National Council on Nutrition and Physical Activity in cooperation with

Institute for Nutrition Research, University of Oslo (NO)- Ministry of Health, National Food and Nutrition Institute, Human Nutrition Faculties on

Universities + under the auspices of EU (PL)- National Institute of Statistics. In local studies, Health Centres may also be important (PT)- Medical Faculty of Comenius University (SK)- EU or Ministry of Health (SP)- NFA, University departments (SW)- Federal Health Institute (SZ)

7. Age limit for signed consent by subject itself (and not parent)?Answers n Countries- 14 years- 15 years- 17 years- 18 years

- No fixed age limit- Unknown, possibly 16- <12: parent; 12-18: child + parent

15112

211

GECZ,DK,FR,PT,SKSWAU,BE,CR,GR,HU,IC,IT,LI,NL,PL,SP,UKFI,SZIRNO

Remarks:- Legally 18 years; Local ethical committees for scientific research advise to obtain the child’s

assent in any case (BE)- 15 is age of criminal responsibility. Between age 15 and 18 the parent has to be informed (DK)- Decided by ethical committee, based on: aim/questions of study, where study is carried out. From

14-15 onwards possible to obtain signed consent only from child. (FI)- 15 years currently used, not known whether there is a general age limit (FR)- 15 years (not official) (PT)- in NDNS: ≤ 15 years: only parent; 16-17 years: both subject and parent; ≥ 18 years: subject (UK)

7

8. Common practice to use incentives or other means?Answers N Countries1 yes

2 noSometimes, dependent on type and extent study

18

32

AU,BE,CR,CZ,DK,FI,FR,GE,IC,IR,LI,NLNO,PT,SK,SW,SZ,UKGR,HU,SPIT,PL

If yes, what is used? (more answers possible)- study results (not specified)- overall study results- individual study results- medical examination- nutritional advice- (small) present- money- lottery- present only for children- depends on survey and amount of money:

- adults and blood samples taken: money £5-£10- children: present (t-shirt, book)

- booklets, folders etc.

5149365611

2

AU,CR,GE,IT,PLBEBE,FI,SK,SWAU,CR,(CZ),GE,IT,(PL),PT,(SW),SZAU,SK,SZ(BE),(CZ),FR,IC,NL,SWCZ,LI,NO,PT,UKCZ,DK,FR,IC,NO,SWGEIR

IT,PLRemarks:- Small present only if the budget allows it and there is a large burden on the respondent (BE)- In HHBS sometimes money. Small present or lottery would be nice (CR)- Usually money or lottery; sometimes medical examination or present (CZ)- Lottery is legally not allowed (IR)- Lottery or check of 100-500 USD, or for children each child a check of 12 USD (NO)- This practice is just starting (book, CD, t-shirt?) (SP)- In NDNS: usually a gift voucher of £5-£10 (UK)If yes, data on effect?Yes, specification:- better accuracy- lottery increased response rate- Mainly in elderly people (money, analysis),

but also in other adults- Gifts small positive effect in household

budget surveys in later 1980s; in our nationaldietary survey response probably influencedby other factors

- Increases commitmentNo data (or not answered)

211

1

115

LI,UKNOPT

SW

UKAU,BE,CR,CZ,DK,FI,FR,GE,IC,IR,IT,NLPL,SK,SZ

Remarks:- Subjective experience: study results and nutrition advice extremely helpful (AU)- Feedback on study results highly appreciated. Modest financial incentives/presents appreciated but

little influence on decision to participate (BE)- Money is the best (CZ)- Getting personal health data is well respected (FI)- Subjective experience: at least about 10 Euro per person needed (FR)- Effect on response unclear, for re-examination of participants it has a positive effect (GE)- L Johansson et al. Eur J Clin Nutr 1997;51:346-53 (NO)- Incentives have to be large to have any marked effect on response. However, small incentives

encourage respondents to provide better quality information and increases commitment (UK)

8

9. Other methods used to handle non-response?Answers N Countries- Questionnaire for reasons of non-response

and some elementary questions (by mail orphone)

- Explain relevance and purpose of study andguarantee confidentiality

- Minimize burden for respondent/flexibleorganisation

- Try to re-interview the non-respondents andif necessary enlarge sample

- Presents- Repeated letters, phone calls or other way of

re-contacting- Home visits with a shorter set of

measurements- Personal communication- Substitution by others with similar

characteristics- Teaching and training of interviewers- Afterwards correction by weighing factors- Use of school and class room setting with

children (>80% response)- individual dietary data- in elderly: arterial tension measure + small

present (chocolates)- Re-issues and postal non-response follow-

upsNone

6

3

2

1

14

1

22

121

11

13

AU,FI,FR,GE,IR,IT

BE,NL,PT

BE,NL

CR

DKFI,FR,IC,IR

FI

GR,NOGR,PL

LINL,SWNO

SKSP

UKCZ,HU,SZ

Remarks:- Sending a letter. The National Institute of Statistics Surveys are compulsory (FR)- Experience over > 10 years is a decrease in response (NL)- Weighting method based on secondary response of sub-sample of primary non-responders (SW)

9

10. Number of times feasible to re-contact persons reluctant to participate or cannot bereached at first time?

Answers n Countries- 1x- 1-2x- 1-3x- 2x- 2-3x- 3x- 4x- 5x- Not reached: 6x; reluctant: 1x- Refusal: not re-contacted or once for ‘non-

responder survey’ (limited number ofcharacteristics to trace selection bias);Cannot be reached: as much as needed; ingeneral 3 attempts

- As much as needed until clear positive ornegative reply

- 2x by telephone (after letter), but depends onmodality of refusal

- Interview: max 3 home visits, max 5 phonecalls; questionnaire: 1-2x reminder

- reluctant: 2x; not reached: 3xNot known

2413121111

2

1

1

11

PL, HULI,NO,SK,SWFIAU,CZ,SPICCR,GEIRDKFRBE

GR,UK

IT

NL

PTSZ

Remarks:- Depending on willingness to respond to an additional short questionnaire (AU)- No limits, tried until the end of the fieldwork period. The minimum number of calls should be 5,

at different times of the day and on different days of the week (UK)

10

FIELD WORK:

11. Age limit to conduct a 24-h recall with a child?Answers N Countries- 6-7 y- 6 or 8-10 y- 7 y- 7-8 y- 8 y- 9-10 y- 10 y- 10-12 y- 11 y- 12 y- 12-14 y- 13-14 y- 15 yNot answered

11211132341111

SPAUDK,LIPLNLNOGR,IC,PTBE,FIGE,IT,SKCR,CZ,HU,SZIRSWFRUK

Remarks:- 6 when conducted by an educated interviewer, otherwise 8-10 (AU)- Supporting information is desirable when children alone are interviewed (e.g. questionnaire on

types of fat/milk etc., list of school lunches) (FI)- 8 years; with assistence of parents (NL)- 13-14 y, but assistance of parents might be needed in some cases (SW)

11

12. Way to carry out an interview with smaller children?Answers N CountriesInterview site:- 4-6 y: preschool- 6-12 y: school or at home- < 7 y: child and maternity centres; ≥ 7 y:

schools; but usually separate dietary studies,together with parents

- at home- usually at school/daycare center- at school- 10-16 y: at school; 6-9 y: at home- small children within the home and school- 3 y: at home; >3 y: at school/day-care centre- restricted information: at school; more

details: at home- 9-12 y: at home; 13+ y: at school- 2-12 y: at home; 13+ y: in school- depends on study type- < 6 y: preschool facilities/kindergartens;

> 6 y: schoolsNot answered

111

2331111

1141

1

AUBEFI

FR,SZGE,GR,PTHU,LI,PLICIRITNL

NOSWCR,CZ,DK,SPSK

UKInterview with whom:- 4-6 y: school staff + parents- 6-12 y: always with help of parents- parent/school staff/daycare staff- child and parent/school staff/daycare staff- child and parent- parent- child only- < 6 y: parent and daycare staff; 6-9 y: parent;

10-16 y: child- <3 y: parents; 3-5 y: parents + school/day-

care staff; > 5 y: school/day-care staff (+possible cross-check by interviewing parents)

- 2-12 y: parent; 13+ y: child aloneNot specified

11335231

1

12

AUBECR,CZ,SKDK,IR,SPFI,FR,GR,NL,SZGE,PTHU,LI,PLIC

IT

SWNO,UK

Remarks:- To interview a child at school, permission of the Ministry of Science and Education is necessary

(LI)

12

13. Feasible to perform interviews on weekend-days and holidays?Category answers N CountriesSaturday a. yes

b. no

16

7

BE,CR,CZ,DK,FI,FR,GR,IC,IR,IT,NL,NO,PT,SK,SZ,UKAU,GE,HU,LI,PL,SP,SW

Sunday a. yesb. no

914

CZ,DK,GR,IC,IR,IT,NO,PT,UKAU,BE,CR,FI,FR,GE,HU,LI,NL,PL,SK,SP,SW,SZ

Holidays a. yesb. noYes, but rarelyNot preferred

101111

BE,CR,CZ,DK,GR,IR,IT,PT,SW,UKAU,FI,FR,GE,HU,IC,LI,NO,SK,SP,SZPLNL

Remarks:- Weekend and holidays are usually seen as very private. Additionally nutrition behaviour is

completely different. (AU)- Only Saturday morning feasible; Holidays: not between 15th July and 15th August (FR)- Feasible, but not always efficient. Degree of urbanisation may influence response rate during

weekends and holidays (GR)- Weekend feasible for interviews at home and perhaps at study site, not at school (IC)- Sundays only in case of unusual circumstances; no interviews in Christmas, new Year and Easter

(IR)- Dependent on available funds (higher costs) + probability of refusal or not reached increases (IT)- Recall on these days of children is not recommended (LI)- Holidays higher non response; more difficulties to organise (NL)- Difficult because interviews are usually performed at schools and working places, sometimes

investigator’s office (PL)- No for children, yes for elderly (SP)- Weekends difficult. Perhaps a telephone interview combined with personal interview on Monday

could be a solution. During holidays more difficult to reach families, but during shorter schoolholidays most parents are at work (SW)

- The only constraint is that interviewer pay is higher for weekend and public holiday working(UK)

13

14. Qualification for interviewer?Answers N Countriesa. dieticians (or nutritionists) (or students)b. home economics (or students)c. other (skilled interviewer in general)b + ca + ca + ba + b + c

11

51321

AU,BE,CR,FI,GE,GR,IR,IT,NL,SP,SZ

CZ,FR,LI,PL,UKDKHU,PT,SKIC,NOSW

Remarks:- Student of nutritional sciences and graduates from nutritional sciences with special education on

the interview technique (AU)- Skilled staff of the hygienic services (dieticians) or skilled students (CZ)- Skilled interviewers, managed by dieticians; no students (FR)- For in-depth food intake surveys dieticians; for telephonic interviews skilled interviewers or

students (IT)- Interviewers are always trained. Many of them have nutritional education or work in the field of

nutrition (PL)- To interview a child at school, permission of the school and regional authority is necessary (SK)- Dependent on study. For 24-hr recall: home economics, dietetics or nutrition preferable (SW)- No formal educational qualifications required for interviewers used in the NDNS (who are

recruited and employed by the survey contractor and work on a range of social surveys).Interviewers must pass a clerical test and are employed on the basis of aptitude and skills trainingand testing. Training is provided on all aspects of interviewing and performance is continuouslyassessed and training given. Interviewers working on the NDNS attend a 4-day training sessioncovering all aspects of the survey. (UK)

14

15. Best structure for European consumption surveillance?Answers N Countriesa. repeated monitoring system

every …. year: 2355-1010

b. rolling systemDepends on kind of data that is needed:- means for trends à rolling system- more detailed data à repeated monitoring

systemNot answered

17

22832

41

1

AU,CR,CZ,FI,FR,GE,GR,HU,IC,IT,LI,NLNO,PL,PT,SK,SPAU,ICCZ,SPCR,FR,GE,GR,LI,NL,NO,SKFI,HU,ITPL,PTBE,DK,IR,SZSW

UKRemarks:- It depends. A rolling system is more systematic approach which needs money as “an overhead” for

some institution. A repeated system can be paid as a discrete, repeated activity. (CZ)- Diet does not seem to change that much in 5 years. A rolling system would need quite a reform at

the national level. (FI)- Subject to funds and resources being available a rolling system is the desired option. This would

allow structures to be put in place in the form of a food consumption surveillance unit. Such asystem would allow amendments to be made in the methodology. Also, expertise and constructiveplanning could be established rather than transient when surveys are repeated periodically.Computer systems and data-entry procedures could be established and in place for the subsequentsurveys. (IR)

- The ideal situation could be to implement an integrated system of data resources: nationwide foodsurveys could provide the information base for planning intermediate in-depth studies, likespecific population groups, target substances, by using different methodologies (e.g. total dietstudies). All gathered information would be used to refine the informative base that necessarilypresents a degree of generalisation that could render it not suitable for specific purposes (e.g.categorisation into food groups could hide food products particularly interesting for certainsubstances or population groups). (IT)

- Using harmonized methodology (PT)

1

Annex E-2

EFCOSUM; WORKING GROUP 4

RESULTS OF THE SECOND QUESTIONNAIRE ONOPERATIONALIZATION OF A EUROPEAN FOOD CONSUMPTIONSURVEY

Participating countries (abbreviated in the tables below):(AU= Austria) GR = Greece PL = PolandBE = Belgium HU = Hungary PT = PortugalCR = Croatia IC = Iceland SK = Slovak RepublicCZ = Czech Republic IR = Ireland SP = SpainDK = Denmark IT = Italy SW = SwedenFI = Finland LI = Lithuania (SZ = Switzerland)FR = France NL = Netherlands UK = United KingdomGE = Germany NO = Norway

(Note: The questionnaire was not available from Austria and Switzerland)

TARGET POPULATION:

1) Most important criterion on individual level for defining the target population?Answers N Countries1 nationality2 chief residence3 language4 religion5 living in country for at least a specified period6 other; specification:

- education and/or occupation, age, gender- private household- specific age range and residence at private

adress7 combinations of the above; specification:

- nationality + language- language + living in country for at least …

1121

111

31

PLCZ,FI,FR,GE,HU,IC,IR,IT,NL,SK,SP,SWCR

LIPTUK

BE,DK,NOGR

2) Different distinct target populations that should be sampled independently?Answers n Countries1 yes2 no

No answer

614

1

FI,FR,GE,GR,IT,NOBE,CR,CZ,DK,HU,IC,IR,LI,NL,PL,PT,SK,SP,UKSW

If yes, specification:- region (= area), 10 year age group, sex- perhaps children in next national individual

survey (2003)- foreigners/emigrants- free-living and institutionalised persons- speaking national language or not (NO)

11

111

FIFR

GEITNO

2

No specification given 1 GR

3

3) Subgroups in the population not belonging to the target population?Answers n Countries1 yes2 no

No answer

812

1

CZ,DK,FI,FR,GE,LI,SP,UKBE,CR,GR,HU,IC,IT,NO,NL,PL,PT,SK,SWIR

If yes, specification:- Gipsydom- immigrants- homeless, addicted- children (separate studies)- hospitals- nursing homes- institutions for elderly- prisons- tourists- illegals- pregnant and lactating women

11112222111

CZDKGEFIFR,LIFR,LIFR,LIFR,LISPSPUK

4) Exclusion criteria?Answers n Countries1 institutionalised persons:

a hospitals

b nursing homes

c institutions for elderly

d prisons

e other institutions; specification:- Army- Therapeutic communes / narcotics- All types of closed catering- Residences for students- Boarding schools, children’s homes etc.

Type of institution not specified2 specific diseases; specification:3 specific handicaps; specification:4 not speaking common language(s) of country5 other; specification:

- pregnant and lactating women- refugees recently arrived- living in country < 1 year

6 no exclusion criteriaNo answer

16

15

14

14

14

411111

7

11121

BE,CR,CZ,DK,FR,GR,HU,IC,IR,NL,NO,PL,PT,SW,UKCR,CZ,DK,FR,GE,GR,HU,IC,IR,NL,NO,PL,PT,SW,UKCR,CZ,DK,FR,GE,GR,HU,IR,NL,NO,PL,PT,SW,UKCR,CZ,DK,FR,GE,GR,HU,IR,NL,NO,PL,PT,SW,UKCR,CZ,DK,FR,GE,GR,HU,IR,NL,NO,PL,PT,SW,UK

HU,NO,NL,PTCRCZPTUKBE

BE,CR,DK,FI,GR,NL,PT

LISWPTIT,SKSP

Remarks:- Conducting a survey on other languages than Finnish and Swedish may be problematic (about

1% of the population) (FI)- France: Same exclusion criteria as mentioned in question 3 (FR)- Specific diseases, physical disabilities, language difficulties not specifically excluded, but they

may be effectively excluded because they are not able to participate (UK)

4

SAMPLING FRAME:

5A) Type of sampling frame preferred?Answers N Countries1 population register

2 census3 electoral list4 list of households5 market research panel6 other; specification:

- postcode adress file – list of addressesTwo or more types indicated:- census + list of households- Master sample (= representative of private

households), got from last census

12

22

11

21

BE,DK,FI,GE,HU,IC,IT,LI,NL,NO,PL,SWFR,GRCZ,IR

UKUK

CR,SPPT

Remarks:- Census is underway this year; results are expected 2001/2002 (CR)- 2nd answer: list of households/phone list; 3rd answer: market research panel (FR)- Although we have good experience with a market research panel for our national food

consumption surveys, for sampling on individual level in stead of household level the panel mightbe too small. (NL)

- Census for Household Budget survey; for individual level: population register, list ofhouseholds or market research panel (SK)

5

5B) Access to the following sampling frames?Category Answers N Countriesa. population register 1 yes

2 nono answer

1236

BE,DK,FI,GE,HU,IC,IT,LI,NL,NO,PL,SWCR,CZ,UKFR,GR,IR,PT,SK,SP

b. census 1 yes2 nono answer

677

BE,CR,FR,GE,SP,SWCZ,HU,IR,IT,NL,PL,UKDK,FI,GR,IC,LI,NO,PT

c. electoral list 1 yes2 nono answer

7410

BE,CR,CZ,IR,IT,SW,UKGE,NL,PL,SPDK,FI,FR,GR,HU,IC,LI,NO,PT,SK

d. list of households 1 yes2 noyes but notworthwileno answer

371

10

CR,GE,SPBE,CZ,IT,NL,PL,SW,UKFI

DK,FR,GR,HU,IC,IR,LI,NO,PT,SKe. market research panel 1 yes

2 noyes but notworthwileno answer

831

9

CR,CZ,FR,IR,IT,NL,SW,UKBE,GE,PLFI

DK,GR,HU,IC,LI,NO,PT,SK,SPf. other; specification:- Master sample (=

representative ofprivate households),got from last census

- postcode address file

1 yes

1 yes

1

1

PT

UKRemarks:- No definite answer can be given. New census will take place March 18, 2001 (GR)- Master sample: National Institute of Statistics (NIS) selects the sample, is involved in data

collection and provides an anonymous database. Negotiations with NIS will be necessary (PT)- Census: it is anonymous (HU)- Access to Census for Household Budget Surveys (SK)- Census: access under special permission (SP)

6

5C) Relation of accessible sampling frame to the target population?Category Answers N Countriesa. population register 1 complete

2 incomplete, specification:3 extra elements; specification:

- institutionalised people andother language (see 4 exclusion)

4 not applicable (not accessible)not answered but not indicated as not accessible

11

11

36

BE,FI,GE,HU,IC,IT,LI,NL,NO,PL,SW

DKDK

CR,CZ,UKFR,GR,IR,PT,SK,SP

b. census 1 complete2 incomplete, specification:

- emmigrants- refugees

3 extra elements; specification:- institutionalised people

4 not applicable (not accessible)not answered but not indicated as not accessible

32

11

11

68

BE,CR,GRSP,SWSPSWFRFRCZ,HU,IT,NL,PL,UKDK,FI,GE,IC,IR,LI,NO,PT

c. electoral list 1 complete2 incomplete, specification:

age groups < 18 yearsnot specified

3 extra elements; specification:- citizens living abroad

4 not applicable (not accessible)not answered but not indicated as not accesible

16

51

11

49

BECR,CZ,HU,IT,IR,UKCR,CZ,HU,IT,IRUKSWSWGE,NL,PL,SPDK,FI,FR,GR,IC,LI,NO,PT,SK

d. list of households 1 complete2 incomplete, specification:3 extra elements; specification:4 not applicable (not accessible)not answered but not indicated as not accessible

2

712

CR,SP

BE,CZ,IT,NL,PL,SW,UKDK,FI,FR,GE,GR,HU,IC,IRLI,NO,PT,SK

e. market research panel 1 complete2 incomplete, specification:

- no specification mentioned3 extra elements; specification:4 not applicable (not accessible)not answered but not indicated as not accessible

31

1

314

CR,FR,NLSWSW

BE,GE,PLCZ,DK,FI,GR,HU,IC,IR,IT,LI,NO,PT,SK,SP,UK

f. other, please specify:-Master sample (see 5A)

- postcode address file

2 incomplete, specification:- institutionalised people

3 extra elements; specification:- addresses which are not private

households (small businesses)

11

11

PTPTUKUK

Remarks:- Electoral list: it’s a voluntary register (individuals sign on voluntarily) (IR)- Market research companies are not regularly involved in these studies. It depends on the aim of

the study and available budget. Therefore, there are not fixed characteristics to be indicated. (IT)- Census for Household Budget Surveys : complete (SK)

7

6) Supervising institute for preferred sampling frame?Category N

(samplingframe)

Answer N Countries

a. population register 12 - National Institute of Statistics- The Danish Ministry of the

Interior's Central Office of CivilRegistration (the CPR-Office)

- Population Register Centre- “Landeseinwohnermeldeämter” =

Federal state population registers(no central institution)

- National Population RegisterOffice (NPRO)

- Icelandic National Registry- Municipality councillorship- Ministry of internal affairs;

National statistical department- Municipalities- Directorate of Taxes, Statistics

Norway- Ministry of Domestic Affairs

and Administration- Taxation authorities

11

11

1

111

11

1

1

BEDK

FIGE

HU

ICITLI

NLNO

PL

SWb. census 4 - State Bureau of Statistics

- National Institute of Statistics(INSEE)

- National Statistical Office- National Institute of Statistics

11

11

CRFR

GRSP

c. electoral list 2 - The Czech Statistical Office- Department of the Environment

11

CZIR

d. list of households 2 - State Bureau of Statistics- National Institute of Statistics

11

CRSP

e. market research panelf. other; specification:- Master sample- postcode address file

211

- National Institute of Statistics- Post Office

11

PTUK

Remarks:- Census for the Household Budget Surveys : The statistical Office of the Slovak Republic (SK)- Croatia and Spain indicated census + list of households as sampling frame (see 5A). Answer

to this question included with both sampling frames. (CR,SP)

8

7a) Inclusion/exclusion criteria for preferred sampling frame?Category N

(samplingframe)

Answer N Countries

a. population registerInclusion:

Exclusion:

12- all citizens/residence

(i.e. permanent address)- persons locally born +

persons requesting fromother municipalities +foreign people aged over 18provided with a validresidence permit

- Chief residence + ableto understand Germanlanguage + non-institutionalised

- -Criteria unknown

9

1

1

1

BE,DK,FI,HU,IC,NL,NO,PL,SWIT

GE

LIb. census

Inclusion:

Exclusion:

No answer

4- permanent residence for at

least 1 year- illegal migrations- tourists

1

111

CR

SPSPFR,GR

c. electoral listInclusion:Exclusion:

2- ≥ 18 years 2 CZ,IR

d. list of householdsInclusion:

Exclusion:

2- permanent residence for at

least 1 year- illegal migrations- tourists

1

11

CR

SPSP

e. market research panelInclusion:Exclusion:

f. other; specification:-Master sample

Inclusion:Exclusion:

- postcode address file

21

1

- private households

No answer

1

1

PT

UKRemarks:- A Finnish municipality of residence is registered for a foreigner who has moved to Finland if he

intends to stay permanently and if he also has a residence permit for at least one year. Citizens ofthe Nordic countries do not need residence permits. (FI)

- Cannot be answered since the recent census is still in progress (GR)- Census for Household Budget Surveys : Criteria for purpose quota sample: social group of household head; net money income; number of children in economical household;

number of members in pensioner’s households; region (SK)- Croatia and Spain indicated census + list of households as sampling frame (see 5A). Answer to

this question is included for both sampling frames (CR,SP)

9

7b) The frequency of update of the preferred sampling frame?Category N

(samplingframe)

Answer N Countries

a. population register 12 1 continuously (at least weekly)

2 discontinuously; interval?Unknown/not answered

9

3

BE,DK,FI,GE,HU,IT,NL,NO,SW

IC,LI,PLb. census 4 1 continuously (at least weekly)

2 discontinuously; interval? - 10 yearNo answer

21

11

FR,SPCRCRGR

c. electoral list 2 1 continuously (at least weekly)2 discontinuously; interval?: - 1 year

- 2 year

211

CZ,IRIRCZ

d. list of households 2 1 continuously (at least weekly)2 discontinuously; interval? - 1 year

11

1

SPCRCR

e. market research panel 1 continuously (at least weekly)2 discontinuously; interval?

f. other; specification:-Master sample

- postcode address file

21

1

1 continuously (at least weekly)2 discontinuously; interval? - not on regular basis

No answer

11

1PTUK

Remarks:- Croatia indicated both census + list of households as sampling frame. Separate answers to this

question are included for both sampling frames. (CR)- Population register: on-line from the local register offices (FI)- Census: until 1999 every 7 years; it will be continuously in the next years (FR)- Cannot be answered since the recent census is still in progress (GR)- Population register: continuously “at least theoretically” (IT)- Census for Household Budget Surveys : discontinuously: yearly (SK)- Spain indicated both census + list of households as sampling frame. Answer to this question is

included for both sampling frames. (SP)

7c) Other relevant comments on the sampling frame?Category N

(samplingframe)

Answer N Countries

a. population register 12 - The CPR is the most up-to-date andcomprehensive register in DK. It is thenational, official register of personsliving in DK.

1 DK

b. censusc. electoral listd. list of householdse. market research panelf. other; specification:

10

8) Routine procedures of quality control in the preferred sampling frame?Category N

(samplingframe)

Answer N Countries

a. population register 12 - see http://www.cpr.dk/welcome.htm- work in NPRO is strictly

regulated by law and other rules- Norway is a relatively small and fairly

stable community and public usuallyloyal to authorities + well-organisedcivil registration system withpossibilities to check the informationagainst other sources (medical records,stock data etc). Therefore quality isconsidered very satisfactory (Technicaldetails may be provided on request)

- Comparing stratification groups withactual data from German CentralStatistic Office

Unknown/not answered

11

1

1

8

DKHU

NO

GE

BE,FI,IC,IT,LI,NL,PL,SW

b. census 4 - supervision in the field (datacollection, questionnaire) + logical/mathematical control + handling of“outliers” + non-response (unit anditem) + standard error

- check answers by interviewers +regional officers at regional andnational level + comparisons withother files (electricity, schools)

Unknown/not answered

1

1

2

CR

FR

GR,SPc. electoral list 2 - No other than the data on the electoral

list being revised annuallyUnknown/not answered

1

1

IR

CZd. list of households 1 Unknown/not answered 1 SPe. market research panelf. other; specification:-Master sample- postcode address file

211

To be answered later onNot answered

11

PTUK

Remarks:- Croatia indicated census + list of households as sampling frame. Answer to this question

included with census only. (CR)- The accuracy and up-to-date level of personal information as well as information on addresses

is about 98% (FI)- Cannot be answered since the recent census is still in progress (GR)- Specific quality control procedures unknown, but because you need a proof of being registered

in a municipality for all kinds of official things (social or study allowances etc.) it is probablycomplete and correct. Only illegals will be missing. (NL)

- Spain indicated census + list of households as sampling frame. “Unknown” included at bothsampling frames (SP)

- Not answered by Slovak Republik (but sampling frame for household budget survey) (SK)

11

9) Single- or multistage sampling?Answers n Countries1 single-stage2 multistage

Depends on data gathering: single-stage for bytelephone; multistage for by personal interview

515

1

DK,FI,HU,LI,SWBE,CR,CZ,FR,GE,GR,IT,IR,NL,NO,PL,PT,SK,SP,UKIC

If multistage, consecutive steps?- 1. Province; 2. Village/town; 3. Individual- 1. Territorial units; 2. Households; 3.

Individuals- 1. Regions; 2. county/places; 3. Persons- 1. Census, random à district; 2. Census

random selection à household; 3.Composition of household, Kish method àindividuals

- 1. Define communities as urban, fishingvillage and rural à communities; 2. randomsample each type of community à categoriesof communities; 3. Random sample ofindividuals à individuals

- 1. Geographical distribution (Federal state –region – community size – electoral district)à electoral district; 2. Age and sexstratification à persons

- 1. Selection of cluster à clusters ofdistrict electoral divisions (further steps notdefined)

- 1. List of all municipalities à specificmunicipalities; 2. Age/gender categories àindividuals according to age and gender

- 1. Main geographical area à regions; 2.Demographical size à municipalities; 3.Municipality’s areas à households/individuals

- 1. Selection of PSUs à municipalities; 2.Selection of sample à individuals

- 1. Area survey units; 2. Individuals- 1. Census section as the country division; 2.

Individuals- 1. Postcode areas; 2. Addresses; 3.

Households; 4. IndividualsSteps not specified

11

11

1

1

1

1

1

1

11

1

3

BECR

CZFR

IC

GE

IR

NL

IT

NO

PLSP

UK

GR,PT,SK

Remarks:- The subsample for the Dietary Survey has been 40% of the sample for Health Monitoring

purposes – it is however included in the same sampling stage (FI)- The sample will be chosen by the NPRO from the sampling frame. The sample should be

statistically representative to the whole population and the inhabitants of administrative units(regions, towns, villages) by gender, age, and domicile. (HU)

- For postal survey or telephone surveys single stage; for visits in respondent homes or localresearch centres multistage (see above) (NO)

12

INELIGIBILITY, RECRUITMENT AND NON RESPONSE:

10) Definition ineligible: individuals who died or moved out of the region under study in theperiod between the drawing of the sample and the actual screening. Agree?

Answers N Countries1 yes

2 noDepends on individuals or addresses as a sample

17

31

BE,CR,CZ,DK,FI,FR,GE,HU,IR,IT,LI,NLNO,PL,PT,SK,SWGR,IC,SPUK

If no, alternative definition:- “Lost” and coding the reason- excluded- inappropriate

111

GRICSP

Remarks:- Agree, except that also people who moved in our sample size are included (DK)- Sample of individuals: If a named individual has moved then efforts are made to trace them; if

they are untraceable then they are coded as non-contacts. Sampled individuals who are dead areineligible.Sample of addresses (like the NDNS): the situation doesn’t arise – if the address is empty it isineligible.Surveys are based on random probability sample design and so substitutions are never made forsampled individuals, households or addresses. (UK)

11) What should be done with the ineligible subjects (as defined in 10)?Answers N Countries1 substitute with newly drawn subjects from the

original sampling frame2 label as “ineligible” and not replace them3 other; specification:

- prepare a substitutions list/oversampling- substitute, using matching criteria

See remarks question 10

9

83

21

1

BE,CZ,FR,GE,HU,IC,IR,PL,SK

DK,CR,FI,LI,NL,NO,SP,SW

IT,PTGRUK

13

12) Definition of subjects who can not be contacted?Answers n Countries1 after sending …… (number of) letters

number: 22 after …… (number of) telephone calls

number: 4-55

3 after …… (number of) home visitsnumber: 4

severalnumber not mentioned

4 other; specification:subjects not reached by the end of thefieldwork period, including those whereit wasn’t possible to determine whetherthey were eligible

5 combinations of the above; specification:- 2 letters + telephone call- 1 letter + 3 telephone calls- 1 letter + 4 telephone calls- 1 letter + 5 telephone calls- 6 telephone calls + 3 home visits- 3 letters + 4 phone calls + 3 home visits- 2 letters + 3 phone calls + 1 home visit- 2 letters + 2 home visits- 1 letter + 3 home visits- 1 letter + 1 phone call + 1 letter- 3 letters, phone calls, home visit- 3 letters + 2 phone calls

11

211

3111

11

111131111111

HUHUNLGRNLCZ,IR,LIIRLICZUKUK

SWCRICITDK,FR,SKNOSPPLPTFIGEBE

Remarks:- Data Inspectorate permits no more than 3 letters; number of phone calls and home visits are

suggested numbers, not regarding cost assessments (NO)- + information from neighbours about subject being away for a long time (PT)

13) Definition of denominator of the participation rate?Answers n Countries1 A : all subjects in the original sample2 B : all subjects in the original sample minus

“ineligibles”3 C : all subjects in the original sample minus

“ineligibles” minus subjects “not able tocontact for other reasons”

313

5

GR,HU,NLCR,CZ,DK,FI,FR,GE,IC,NO,PL,PT,SK,SW,UKBE,IR,IT,LI,SP

Remarks:- In some studies also the definition C has been included in the reporting in addition to the

definition B (FI)- Definition A + specify non-response: % of ineligibles + % of not able to contact + % of refusals

(NL)

14

14) Recruitment procedures?Answers n Countries- By letter only- Letter + home visit- Involving mass media- Letter or phone call- Letter + phone call- Letter + phone call + visit (repeated

during the survey period, e.g. 3x per week)- Letter + phone call + visit or personal contact

at centre- Letter (+ phone call) + home visit- Letter including consent form and

suggested date for the visit- Letter; if no response a phone call +

another letter- Combinations of letter, phone call, visit- Letter to household followed by home visit;

random selection of individual by intervieweron doorstep

- Letter (+ 2 reminding letters) + personalcontact by telephone or home visit

141161

1

11

1

11

1

HUFR,IR,PT,SPLISKBE,CR,DK,IC,NL,NOIT

SW

CZPL

FI

GRUK

GE

Remarks:- Confirmation of participation during 1 to 1 meeting (IR)- 1. Letter with information about the study + possibility to refuse to be contacted further (by

telephone number or reply card); 2. Telephone call for further explanation and ask for participation(if no telephone number is available by home visit) (NL)

- Letter is mandatory; telephone call is preferred due to cost considerations (NO)- No representative national nutritional survey conducted so far (SK)

15a) Legal or ethical regulations with regard to number of times you can contactindividuals?

Answers n Countries1 yes2 no

No answer

218

1

IC,NOBE,CR,CZ,DK,FI,FR,GE,HU,IR,IT,LI,NLPL,PT,SK,SP,SW,UKGR

If yes, specification:- Ethical committee must give permission for

procedure- Surveys are voluntary, refusals may guard

against reminders

1

1

IC

NO

Remarks:- The Greek Statistical Office can provide us with the information (GR)- No general regulations. But ethics committees may impose limits/conditions for particular

surveys (UK)

15

15b) Legal or ethical regulations with regard to the way you contact individuals?Answers N Countries1 yes2 no

No answer

614

1

GE,IC,NL,NO,PL,SWBE,CR,CZ,DK,FI,FR,HU,IR,IT,LI,PT,SK,SP,UKGR

If yes, specification:- Ethical committee must give permission for

procedure- The information about the study must be

given on paper; the subject must be given theopportunity to refuse to be further contactedafter the first letter; you may not put pressureon the subject to participate; they must begiven enough time to decide whether theywant to participate

- Personal if possible- Home visit following a reply of the

participant for the letter sent before- The study design generally has to be

approved by an ethical committee. Therespondent must be informed in a properway, i.e. on options to decline etc.

No specification given

1

1

11

1

1

IC

NL

NOPL

SW

GERemarks:- The Greek Statistical Office can provide us with the information (GR)- No general regulations. But ethics committees may impose limits/conditions for particular surveys

(UK)

16

15c) Legal or ethical regulations with regard to the type of feedback to participants?Answers N Countries1 yes2 no

No answer

812

1

FI,HU,IC,IT,NO,PL,SW,UKBE,CR,CZ,DK,FR,GE,IR,LI,NL,PT,SK,SPGR

If yes, specification:- Ethical committee must give permission

for procedure- In case health problems would be found

during the survey, the subject would be askedto contact the health care professionals for afurther health check.

- Through local pratitioner if any factrelating to health (ethical obligation)

- Only upon stated consent- No money, no gifts- Possibility of the contact with the

principal regional investigator(name/address/phone)

- There are certain upper limits for the value ofgifts/lottery prices in case of a governmentalagency, university institution etc. acting asthe principal investigator.

- Feedback is subject to a code ofconfidentiality. It can only be given to therespondent, unless consent is specially givenfor feedback to be given to a third party (e.g.blood pressure results to a subject’s GP)

1

1

1

111

1

1

IC

FI

HU

NOITPL

SW

UK

Remarks:- The Greek Statistical Office can provide us with the information (GR)- You have to do what you promised at the start; participants must be given the opportunity to

refuse to get results from for instance blood values. (NL)

17

15d) Legal or ethical regulations with regard to type of information in the informed consent?Answers N Countries1 yes

2 noNo answer

15

42

BE,CZ,DK,FI,GE,HU,IC,IT,NL,NO,PL,SK,SP,SW,UKCR,IR,LI,PTFR,GR

If yes, specification:- Identification data for person- Guarantee anonymity/confidentially- Aim of the study- Letter with information about aim,

nature, methods and duration of study,burden to subject and possible risks.Declaration that subject is informed.

- Participation voluntarily/withdrawalpossible

- Data only used for purpose of the study- Possibility that data could be used in

different statistical studies (if applicable)- Ask for permission to save blood for

other measurements- Period for which individual data is kept

identifiable- All registers and variables that are

merged to the individual records- Rights of the participant- Give enough time to participant to

understand and give informed consent- Political, religion- See also 15b- Ethical committee must give permission for

procedure

1543

411

1

1

1

11

112

CZBE,DK,GE,NL,PLGE,HU,IT,PLFI,NL,SK

FI,NL,PL,UKPLIT

NL

NO

NO

FIFI,NL

SPSWIC,SK

Remarks:- The Greek Statistical Office can provide us with the information (GR)- According to deontological agreement (281/99) (IT)- Consent form includes name of study and responsible investigator. It must be dated and signed

by both participant and investigator. (NL)

18

15e) Legal or ethical regulations with regard to other recruitment aspects?Answers N Countries1 yes2 no (or not answered)

813

CR,FI,FR,PL,PT,SK,SW,UKBE,CZ,DK,GE,GR,NO,HU,IC,IR,IT,LI,NL,SP

If yes, specification:- Children < 18: parents consent needed- Gifts/payment for participants- No specification givenSee remarks

2222

CR,SKFI,PLPT,SWFR,UK

Remarks:- You have to send a presentation of the survey to a commission (CNIL) and they can ask you to

change the protocol on a case by case basis (FR)- The Greek Statistical Office can provide us with the information (GR)- Household Budget Surveys : internal instructions of the Statistical Office (SK)- Information gathered on subjects is covered by the Data Protection Act. Individuals have access to

information they provided or is held about them as an individual (UK)

19

16) Participation rates in different age and sex strata?Answers N Countries- overall 36%; lower in higher age classes- stable panel of households 100%; other

studies 40-60%- 65% in first Croatian health survey- 1997 Dietary Survey of Finnish adults:

men 60-78%, women 72-79% (see annex 1for detailed table)

- Lower participation among elderly > 75years

- Men 56-71%, women 50-69% (see annex1 for details)

- For household budget surveys only data ontotal participation rate

- men 62%, women 65% in recent IUNAsurvey (North-South Food Cons Survey)

- INN-CA Study 1994-96: 47% ofhouseholds. Highest non response in thenorthern-eastern geographical area.

- men 53%, women 55%; age groups 40-60% in Norwegian survey 1997 (see annex 1for detailed table)

- men 62-74%, women 66-74% +differences in age groups in third DutchNational Food Consumption Survey 1997/98(see detailed table in annex 1)

- Household budget survey 1999: 11-40%(see annex 1 for details)

- About 80%- Household budget survey: 58% (SP)- men 57% (52-68%), women 60% (52-65%)

in Swedish national dietary survey 1997/98(see annex 1 for detailed table)

- Dietary surveys of individuals with dietaryrecords: 11/2 – 41/2 years 80%, 4-18 years64%, ≥ 65 years 59%

- 30-50% in individual nutritional surveys;better response rates in females and highersocio-economic classes.

- Data available but not provided (yet) No data

11

11

11

1

1

1

1

1

1

111

1

1

13

BECZ

CRFI

FRGE

GR

IR

IT

NO

NL

PL

PTSPSW

UK

SK

DKHU,IC,LI

Remarks:- Hungary provided data with only absolute numbers (HU)- Only comparison between age distribution in national registry and in participants (IC)

20

FIELD WORK:

17A) Feasible to do 24-hour recall interviews by telephone?Answers n Countries1 yes

2 noDo not know

14

52

BE,CR,CZ,DK,FR,GE,IC,IR,IT,NO,NL,SK,SP,SWFI,GR,HU,LI,PLPT,UK

Remarks:- In principle feasible, but we do not feel that a 24 hour recall is sufficient to cover our needs in a

nationwide survey, and the participation rate will drop (DK)- Finnish public not happy about telephone interviews; no hard data (FI)- Feasible, but not preferred (GE)- Low coverage of telephones and disproportion between urban and rural areas (PL)- Possibly – untested. Repeat interviews with some face to face and some by phone may be feasible

(UK)

21

17B) Percentage coverage for fixed telephones among households?Answers n Countries- 22% (data 1999 % of inhabitants)- 31% (data 1999 % of inhabitants)- 70%- 76% (% of dwellings; 36% of inhabitants)- 77%- 75-80% (1994/95)- 85%- 90%- 94%- 95%- > 95%- 96%- 98%- about 100%No data or not answered

111111111121224

PLLICZHUFIPTIRCRGENOBE,ICFRNL,SWIT,SPDK,GR,SK,UK

Different across social classes?- Yes, difference (see specification)- No difference- No data, but (probably) different- No data, but probably not (much) different- No data or not answered

124311

NLFR,IRLI,PL,PT,SPDK,FI,NOBE,CR,CZ,GE,GR,HU,IC,IT,SK,SW,UK

Specification difference:- with decreasing SES-level in 5 categories: 100%, 99%, 98%, 96%, 97%. However differences in

secret numbers higher: respectively: 8%, 12%, 14%, 19%, 19%. (NL)Remarks:- Age differences: over 70% of young people (students and small households age < 30) only have

mobile phones. Fixed phones are decreasing and mobile phones are increasing. (FI)- Declining % because of mobiles (FR)- Coverage equal in East and West Germany (GE)- Regional differences in Hungary: minimum 64% Southern Great Plan, maximum 88% Central

Hungary incl. Budapest (HU)- Lowest coverage in age cohort 20-29 (NO)- Large differences between urban and rurals (large rural population: 38% of total) (PL)- Slovak Republic: fixed phones in 1998 45% but now much higher; new data will be available end

of May 2001, also on differences across social classes (SK)- Almost 100% including fixed, mobile and neighbours (SP)

22

17C) Percentage coverage for mobile telephones among individuals?Answers N Countries- 9% (data 1999)- 10% (data 1999)- 20%- 22% (households)- 30%- 35% (households)- 50%- 56%- 73%- 76%- 80%No data or not answered

111111211119

LIPLCZFRHUITCR,GENLNOICFIBE,DK,GR,IR,PT,SK,SP,SW,UK

Different across social classes?- Yes + specification:

- 1998: highest among graduated, males,employed, 25-44 years old

- 16% among household head manualworker, 35% among household headexecutive manager

- higher among: singles, male, 20-30 y, higheducated, lower salary

- No difference- No data, but probably different- No data, but probably not (much) different- No data or not answered

31

1

1

4716

FR,GE,ITIT

FR

GE

DK,IC,LI,SKCR,HU,IR,NO,PL,SP,SWFIBE,CZ,GR,NL,PT,UK

Different across regions?- Yes + specification:

- 1998: centre 42%, north 36%, south 31%- No difference- No data, but probably different- No data, but probably not different- No data or not answered

11

2

216

ITITIC,FI

DK,LIBE,CR,CZ,FR,GE,GR,HU,IR,NL,NO,PL,PT,SK,SP,SW,UK

Remarks:- It differs considerably between age groups (DK)- Age differences: 90% of population 15-40 years of age has mobile phones; lower among elderly

(FI)- 56% business excluded; 62% business included (NL)- 73% of population ≥ 10 years. Best coverage among teenagers, decreasing by age (NO)- Data of 1999; rapid grow in recent years. Different according to occupation and income level.

Highest users among businessmen with high incomes (PL)- Probably high coverage of mobiles. Data available May 2001 (PT)

23

17D) Accessibility of telephone numbers?Category Answers N Countriesa. fixed phones - 75%

- 80%- 81%- 84%- 80-90%- 90%- 99%- (almost) 100%- numbers registered, but

% secret numbersunknown

- numbers registered;nothing mentioned aboutsecret numbers

No data or not answered

111111112

3

8

GENOFRNLBESWCRHUIR,SK

CZ,IT,PT

DK,FI,GR,IC,LI,PL,SP,UK

b. mobile phones - not accessible- 9%- partly accessible- 70%- no secret numbers (but

registered?)- numbers registered, but

% secret numbersunknown

- in preparation (July2001)

No data or not answered

21211

2

110

CZ,NOGENL,PTCRHU

IR,SK

ITBE,DK,GR,IC,FI,LI,PL,SP,SW,UK

Remarks:- Fixed phones: 15% secret numbers (and 96% fixed phones) (FR)- 2000: 4,1 % have exclusively mobile; 97,1 % has either fixed or mobile telephone; Secret

numbers can be accessed with complex procedure of random digit dialing. (GE)- Fixed phones: 14% secret numbers; difference in SES: 8%, 12%, 14%, 19%, 19% with

decreasing SES; mobile phones low accessibility (NL)- Accessible data bases. Automatic search identifies telephone numbers for approximately 80% of

a random sample of adults (NO)

18) Most feasible location for 24-hour recall by personal interview?Answers N Countries1 home visits at the participants home

2 visit of the participant to local research centres3 travelling research centre (research caravans)4 other; specification:- both 1 and 3- both 1 and 2

13

6

11

BE,CR,CZ,DK,FR,IR,IT,LI,NL,PL,PT,SK,UKFI,GE,HU,IC,NO,SW

SPGR

Remarks:- Consulting room of local practitioner (HU)- In general local research centres, but for some persons visit at home (SW)

24

19A) Feasible to include 24-hour urine collection?Answers N Countries1 yes2 noOnly in a subsampleUnknown

10461

BE,CR,GE,GR,IR,IT,NL,NO,SP,UKFR,HU,PT,SKCZ,DK,FI,IC,PL,SWLI

If no, specification:- No infrastructure for sample collection and

transport- Hard to get collaboration for that

1

1

HU

PTRemarks:- Only in a subsample because too costly and would affect participation (IC)- Logistics too cumbersome (for total population) (DK)- Too complicated for the subject; a pilot study is in course (FR)- In a sub-sample of the population, only, with access to regional centres (SW)- Currently collected in NDNS of adults 19-64 years. Unlikely to be feasible in children. (UK)

19B) Feasible to include blood sampling?Answers n Countries1 yes

2 noOnly in a subsampleUnknown

15

141

BE,CR,DK,FI,FR,GE,GR,HU,IR,IT,NL,NO,PT,SP,UKSKCZ,IC,PL,SWLI

If no, specification:

Remarks:- yes, but quite difficult (CR)- Only in a subsample because too costly and would affect participation (IC)- In a sub-sample of the population, only, with access to regional centres (SW)

25

19C) Reduction of participation rates by 24-hour urine collection and blood sampling?Category Answers N Countriesa. 24-hour urine collection 1 yes

2 no(not feasible)

18

21

BE,CZ,DK,FI,FR,GE,GR,IC,IR,IT,LI,NL,NO,PL,PT,SK,SP,SWCR,UKHU

If yes, specification:- More inconvenience to participant (FI)- No data, but 24 hr urine probably unacceptable to many (IC)- Difficult to collect (NO)- In view of working/school hours and time for other activities (PL)- Children and active populations (SP)b. blood sampling 1 yes

2 no

15

6

BE,CR,CZ,DK,FR,GR,IR,IT,LI,NL,PL,PT,SK,SP,SWFI,GE,HU,IC,NO,UK

If yes, specification:- If blood sampling is obligatory some people are not willing; only blood drop is easier (CR)- Involvement of Health Authorities to assure safety and appropriate information campaign would

be needed (PT)- Children and active populations (SP)Remarks:- There is a big resistance in our population after census now (CZ)- No reduction of participation rates as long as participants can take part in some components but

not others if they wish (UK)

19D) Preferred approach for inclusion of 24-hour urine collection and blood sampling?Answers N Countries1 Inform on all aspects and agreement to

participate includes all aspects together.2 Inform on all aspects, but agreement to

participate can be given to the different aspectsseparately.

3 Biomarkers as a second phase of the study;information and ask for participation after thefood consumption part.

4 Other, please specify:

12

9

BE,CR,DK,FI,GE,IC,IR,NL,SK,SP,SW,UK

CZ,FR,GR,HU,IT,LI,NO,PL,PT

Remarks:- Subjects must be informed at the start, but permission to participate can be asked after the food

consumption part (also separate informed consent form) (NL)

26

19E) Regulations about who is allowed to take blood samples? (lowest qualification mentioned)Answers n Countries1 doctor/physician2 qualified nurse, only under supervision of a

physician3 qualified nurse4 trained field worker, only under supervision of

a physician5 trained field worker6 other; specification:- qualified laboratory technician- anyone with phlebotomy training- no specific regulations (see remarks)

23

72

3

211

GR,SKCR,FR,GE

CZ,HU,IT,LI,PL,SP,SWBE,DK

IR,NL,PT

IC,NOUKFI

Remarks:- If only blood drop: qualified nurse without supervision (CR)- No official regulations; study must be carried out under supervision of a physician and worker

needs to be trained under his supervision. Responsible investigator is responsible to ethicalcommittee. Usually qualified nurses or laboratory personnel take blood samples (FI)

- No specific rules; qualified field workers with course on clinical analysis (PT)- Ethical committees need to be satisfied that blood takers are adequately trained. In NDNS fully

trained experienced phlebotomists are used (UK)

19F) Most feasible location to take blood samples?Answers N Countries1 home visits at the participants home2 visit of the participant to local research centres3 travelling research centre (research caravans)4 other; specification:- both 2 and 3- Area Health Authority

4111

41

IR,IT,LI,UKBE,CZ,DK,FI,GE,GR,HU,IC,PL,PT,SWSK

CR,FR,NO,SPNL

Remarks:- Consulting room of local practitioner (HU)- namely Health Centres (PT)

27

20) Picture book with food portion sizes available?Answers N Countries1 yes, Epic picture book2 yes, another national picture book3 noBoth Epic + another national picture book

41241

GE,GR,NL,NOBE,CR,FI,FR,IC,IR,IT,PL,PT,SP,SW,UKCZ,HU,LI,SKDK

If 2, another national picture book:How many different foods are included?

- 24- 24 food items or categories- 25- 65- 78- 80- 107- 126 + 20 new in 2001- 150- about 160 foods (13 image series)- 200- 201- not answered

1111111111111

ICSWCRIRUKPTITFIBEDKFRPLSP

If 2, another national picture book:How many different portion sizes perfood are included?- 1 to 3- 2 to 4- 2 to 6- 3- 3 to 4- 3 to 7- 4- 4 to 6- 5- 8- not answered

11122111111

ITBEIRCR,PLFI,PTSWICDKFRUKSP

Remarks:- Picture book in preparation. It will start with 100 foods and intention is 3 portion sizes per food

(CZ)- Refs: Haapa E et al. Annoskuvakirja. National Public Health Institute, University of Helsinki,

1985. Validation study by Pietinen et al. 1988. (FI)- Book has to be revised (IC)- It is the most commonly used foods atlas (Istituto Scotti Bassani, Milan), but sometimes other self-

developed food pictures folders are used (IT)- Fruit 3 portion sizes in real dimension; other foods 4 in reduced size (PT)- Slovak Republic: there is a picture book used by authors of specific studies; number of foods

depends on the questionnaire (SK)

28

21) List or book with standardized weights or measurements of foods available?Answers N Countries1 yes

2 noIn preparation

15

51

CR,DK,FI,FR,GE,GR,IR,NL,NO,PL,PT,SK,SP,SW,UKCZ,HU,IC,IT,LIBE

If yes, reference:- could copy it (CR)- Mad, mål og vægt (Inge Hansen, National Food Agency ) cited in Scandinavian Journal of

Nutrition, 4/1996 + own lists of registrations based on weighed records (DK)- Lallukka and Ovaskainen: Ruokamittoja. Publications of the National Public Health Institute

B3/2001, Helsinki 2001, 37 pages.(in Finnish). (FI)- Included in the DISHES Program; others are available (MONICA,BLS) (GE)- IUNA food portion sizes (not published) + Helen Cawley Food Portion Sizes (IR)- Rapport Maten Gewichten en Codenummers 1997, Vakgroep Humane Voeding Wageningen,

TNO Voeding Zeist (NL)- Blaker B and Aarsland M. Household measures and weights of foods (in Norwegian). National

Association for Nutrition and Health, Otta 1989. (NO)- List developed by National Food and Nutrition Institute (PL)- Slovak Food Data Bank (as a part of the database of physical properties) (SK)- Vikttabeller (Weight tables, NFA 1999) (SW)- Ministry of Agriculture, Fisheries and Food. Food Portion Sizes, 2nd Edition, The Stationery

Office 1993 (UK)To be delivered at next meeting (PT)Not answered (FR,GR, SP)Remarks:- Only limited list of standards; book in preparation (CZ)

29

22A) Differences in food consumption between week- and weekend days?Answers N Countries1 yes, data available and differences

2 no data available or not answered

13

8

DK,FI,FR,GE,GR,HU,IR,IT,NL,SK,SP,SW,UKBE,CR,CZ,IC,NO,LI,PL,PT

Specifications:- Data will be sent (DK)- 1997 Dietary Survey of Finnish Adults: mainly higher % of energy from alcohol on Saturdays

(FI)- Data not published (FR)- Unpublished results of validation studies (GE)- No data given (GR)- First Hungarian Representative Nutrition Survey (1985-1988): Average energy intake on Sundays

300 kcal (about 10-15%) higher than on working days (Monday – Saturday) (population 15 yearsand older (See Ernährungsforschung 1993;38: 239-246 + reports) (HU)

- First Dutch national food consumption survey 1987/88: Weekend days compared to weekdays:higher absolute intake of most nutrients and energy, lower nutrient density (see detailed results inannex 2) (NL)

- Unpublished data INN-CA Study 1994-96 about differences at food level (see table in annex 2)(IT)

- Preliminary data: higher energy intakes on Fridays and Saturdays compared to other days (IR)- Only old data from 1980’s (NO)- Data of survey of 2000 are now being processed (PL)- Probably Sunday different, Saturday no large differences (PT)- Differences between Sundays and other days (but data not published) (SK)- In weekend more breakfast, during weekdays more eating out (SP)- Data sent to Boeing/Hoffman for analysis (SW)- Differences between week- and weekend days and between Saturdays and Sundays:

- Hughes JM et al. The British National Diet and Nutrition Survey of people aged 65 years andover: protocol and feasibility study. Proc Nutr Soc 1995;54:641-643;

- Maisey S et al. Variation in food group and nutrient intake with day of the week in an elderlypopulation. Br J Nutr 1995;73:359-373 (UK)

30

22B) Experience with 24-hour recalls with respect to the day before?Answers n Countries1 yes, experience2 no experience

714

FI,GR,HU,NO,PL,PT,SKBE,CR,CZ,DK,FR,GE,IC,IR,IT,LI,NL,SP,SW,UK

If yes, specification:- The field workers of the FINDIET 1997 Study considered the 48h-recall a real challenge, but can

be managed. 48h-recall has also been used in Finland in the Multi Center Study of Children(professor Leena Räsänen). (FI)

- No data given (GR)- By means of combination of dietary recall with simplified dietary record (HU)- 48-h recall by Kari Solvoll (copies will be sent) (NO)- Little experience with small survey of 38 year old males: Interviews on Monday with respect to

Saturdays and Sundays (PL)- Not studied, but as a rule when yesterday food intake was not typical the day before is asked for

(PT)- This method has been used in majority of nutrition-related studies in Slovakia in recent 3 decades

(SK)

Other relevant remarks with respect to operationalization:

- It is important the definition of technical aspects for implementing a European monitoring systemof food consumption aimed at analysing nutritional issues, regulatory aspects will be defined inorder to 1) push governmental actions in the various Countries, 2) establish the co-ordinating centre(IT)

31

ad ANNEX E-2: Question 16: Detailed tables of participation rates according to sex and age inlarge population based epidemiological studies:

Finnish Dietary survey of adults of 1997

Age groups Participation rate (%)Men

Participation rate (%)Women

25-34 60 7235-44 68 7445-54 70 7755-64 78 79Total 69 76

German Survey

years 18-19 20-29 30-39 40-49 50-59 60-69 70-79Men (%) 71.4 56.5 61.4 60.6 64.7 63.9 67.7Women (%) 68.7 59.4 65.1 65.5 65.0 61.3 50.4

Netherlands survey of 1997/98

Participation rates third Dutch Food Consumption Survey 1997/98, according to age and gender (Note:non response to the consumer panel is not taken into account)

Age groups Participation rate (%)Men

Participation rate (%)Women

1-4 73.4 69.64-7 69.3 73.07-10 65.0 74.010-13 67.9 70.113-16 64.6 66.516-19 65.4 65.619-22 61.6 57.422-50 65.9 68.850-65 73.7 74.265+ 66.8 69.0Total 68.5%

32

Norwegian surveys of 1993-94 + 1997:

Response rate (%) by gender, age, residence and education in the Norwegian national dietary surveys(Norkost) 1993-94 and 1997.

Response rate (%)

Total number (n)1993-94(3144)

1997(2672)

GenderMen 62 53Women 64 55Age (year)16-19 70 6020-29 64 5430-39 66 5540-49 66 5850-59 66 5660-69 61 5170-79 46 40Part of NorwayCapital and surroundings 60 50East 63 53South 66 59West 65 55Middle 65 54North 62 57Education (years in school)Short (<10) 52 47Medium (10-12) 66 59Long (>13) 75 70

33

Swedish Survey

Participation by sex and age in Swedish nationaldietary survey “Riksmaten 1997-98”

No insample

Sex TOTAL

Male FemaleAge18-24 54% 59% 56%25-34 55% 64% 59%35-44 63% 67% 65%45-54 52% 69% 61%55-64 52% 52% 52%65-74 68% 54% 61%TOTAL 57% 62% 60%

Polish Household Budget Survey

In Polish household budget surveys the sampling unit is a household and information on participation/refusal rates are available for the specific socio-economic household groups. There are seven suchgroups. In 1999, non-participation rates for those groups were: households of retired persons-40.5%,employees’ households-32.4%, households of the persons receiving disability benefits-29.6%,households of the self-employed-24.9% and households of persons living on unearned sources-22,6%.Lower non-response rates were noted for farmers’ households (16,5%) and employees/farmers’households (11.3%)

34

ad ANNEX E-2: Question 22A: Detailed results on difference between week- and workingdays

Netherlands survey

Data of First Dutch National food consumption survey 1987/88:Weekend days compared to weekdays:- Higher absolute intake of: energy, fat (total, saturated, monounsaturated,

polyunsaturated), cholesterol, carbohydrates (especially mono- and disaccharide),alcohol

- Lower absolute intake of: dietary fibre, vitamin B1, vitamin B6- Higher En% of alcohol- Lower En% of protein- Lower vitamin B6/protein ratio- Lower intake of vitamins, minerals and dietary fibre per MJ(Int. J of Food Sciences and Nutrition, 1994, volume 45, supplement 1)

35

Italian Survey

Comparison of food intakes between working days and Sunday or bank holidays in Italy.Data bank of the INN-CA Study 1994-96

Total sample Working days/Sunday-bank holidaysCODE FOOD GROUP Mean Median St.Dev. %consumers

1Cereals and products > < > >2Vegetables and legumes > < > >3Fruit and fruit juices > < > >4Meat < < < >5Fish and seafood > < > >6Eggs > < > >7Milk and yoghurt > < > >8Cheese > < > >9Cream < < > >

10Oil and fats > < > >11Sugar and sweet products < < > >12Beverages > < > >13Special products (baby foods,

dietetic foods, etc.)> < > >

14Ready-to-eat dishes < < > >

Consumers only Working days/Sunday-bank holidaysCODE FOOD GROUP Mean Median St.Dev.

1Cereals and products < < <2Vegetables and legumes < < >3Fruit and fruit juices < < >4Meat < < <5Fish and seafood < < <6Eggs < < <7Milk and yoghurt < < <8Cheese < < <9Cream < < <

10Oil and fats < < >11Sugar and sweet products < < <12Beverages < < >13Special products (baby foods,

dietetic foods, etc.)< < >

14Ready-to-eat dishes < < <Source: National Research Institute for Food and Nutrition, Via Ardeatina 546, 00178 Rome, Italy

Annex E-3

Main structure and principles of standardization of the 24-hour recall performed withEPIC-SOFT

EPIC-SOFT is a software program to perform 24-hour recalls developed within the EPICstudy with the aim to ensure the highest possible level of standardization within and betweena large number of European centres, and to prevent or minimize potential sources of error(Slimani et al., 1999 and 2000). The general structure of EPIC-SOFT is designed to guide andcontrol the interviewer by standardizing the pathways to be followed during the interview. Aflow diagram of the EPIC-SOFT interview procedure is presented in the figure on thefollowing page (Slimani et al., 1999). The main structure and principles of standardization aredescribed below.

To reduce error due to memory two main phases are distinguished. The first phase is a quicklist. This list is divided in common standardized chronological food consumption occasions,for which questions on place and time are asked as a memory aid. Then for each consumptionoccasion, the subject is asked to name all the foods that were consumed. During this stage, thesubject is not interrupted in order to avoid loss of concentration and interruption in memoryretrieval. Per eating occasion there is a country specific check list to ensure nothing isforgotten and at the end a summary of the individual consumption is presented on screen tocheck on completeness.

In the second phase, additional questions are asked to identify, describe and quantify all foodsmentioned in the quick list. For this phase a pre-entered country-specific food list is used.Most rules to describe, quantify and check pre-entered foods and recipes are automaticallyprompted during the interview. This prevents differences between interviewers and ensuresstandardization in level of detail. For quantification different methods are used (photographsof household measures, standard units, food-specific photographs, known amount in weight orvolume, standard portions). The methods that are offered by the system depend on the typeand state of the food (powder or liquid, raw or cooked, with or without inedible parts) and aretherefore also standardized between interviewers. All recipes are broken down into simpleingredients. The total amount of the recipe consumed is estimated, and all ingredients areidentified and quantified in the same way as single foods. All foods and ingredients arecalculated as finally consumed. Algorithms and standard food-specific coefficients forcorrection for inedible parts and/or conversion from raw to cooked and defaults for unknownportion sizes are pre-entered in the system and can easily be updated.

All information is interactively entered by the interviewer and automatically coded.Automatically, checks on completeness and consistency (outlier values, empty fields,incomplete food description and quantification) are performed during the interview, whichallows the possibility to solve these problems while the subject is still present. In this wayarbitrary decisions on outlier values or unlikely food data during data processing areprevented as much as possible. At the end of the interview, as a last check, energy andmacronutrient intake is calculated and compared to standard energy requirement (based onage, sex, weight and height). In this way errors and differences in competence, backgroundand dietary knowledge among interviewers are minimized.

Rep

rint

ed f

rom

‘C

ompu

ter

Met

hods

and

Pro

gram

s in

Bio

med

icin

e’ (

see

refe

renc

es)

with

per

mis

sion

fro

mE

lsev

ier S

cien

ce

References:

• Slimani N, Deharveng G, Charrondiere RU, van Kappel AL, Ocké MC, Welch A, LagiouA, van Liere M, Agudo A, Pala V, Brandstetter B, Andren C, Stripp C, van Staveren WA,Riboli E. Structure of the standardized computerized 24-h diet recall interview used asreference method in the 22 centers participating in the EPIC project. European ProspectiveInvestigation into Cancer and Nutrition. Comput Methods Programs Biomed 1999;58:251-66.

• Slimani N, Ferrari P, Ocké MC, Welch A, Boeing H, van Liere M, Pala V, Amiano P,Lagiou A, Mattisson I, Stripp C, Engeset D, Charrondière RU, Buzzard M, van StaverenWA, Riboli E. Standardization of the 24-hour diet recall calibration method used in theEuropean prospective investigation into cancer and nutrition (EPIC): general concepts andpreliminary results. Eur J Clin Nutr 2000;54: 900-17.

This report was produced by a contractor for Health & Consumer Protection Directorate General and represents the views of thecontractor or author. These views have not been adopted or in any way approved by the Commission and do not necessarilyrepresent the view of the Commission or the Directorate General for Health and Consumer Protection. The EuropeanCommission does not guarantee the accuracy of the data included in this study, nor does it accept responsibility for any use madethereof.